Closed inexxt closed 3 years ago
Ok, I've been thinking about this.
Fin n
, i.e. normalize first.Fin n
, also seems doableThe road to this seems to be best done via having a normal form for Pi combinators (i.e. exactly what @sabry is currently working on).
I still don't see why transpositions would be any easier/harder than Lehmer codes (say). The hard part will still be the equivalent to item 4 above.
As far as I understand, @sabry is trying to express a combinator in terms of transpositions, not to calculate its normal form - this is exactly what the Coxter/
part of the proof is supposed to do.
The code that I've seen being checked-in says otherwise!
Almost solved, the remaining case is going from Pi to Pi^ and back. A separate issue has been created for that #77
Just to 'ack' that I've seen this, and ramping up to be able to work on it.